Local circuit neurons are major elements in the integrative processes in the brain that transform a sensory signal into an appropriate change in motor behaviour. By choosing the most advantageous nervous system for study, the integrative mechanisms in these neurons can be analysed, and the processing followed through all its various steps. The neurons can be treated as identified individuals so that information on their cellular, synaptic and network properties can be collated over the course of many experiments. This proposal has the long-term objective of understanding the design principles for the functioning of local circuits that control locomotion, by studying the networks that regulate leg movements in an insect. It has two specific objectives. First, to analyse the feedback mechanisms that occur within the local circuits particularly at the first stages in the processing. The terminals of sensory neurons from proprioceptors receive a synaptic input from other sensory neurons of the same receptor that reduces the effectiveness of their outputs to other neurons. The hypothesis to be tested is that these interactions form a gain control mechanism that regulates the action of one sensory neuron in the context of others from the same organ. Second, to analyse the processing of the different parameters of the sensory signals coding a joint movement in the terminals of the sensory neurons, in the spiking and nonspiking neurons and in the motor neurons. To achieve these objectives physiological methods of intracellular recording will be used to determine the membrane properties of selected neurons and the patterns of their connections. These will be supplemented where appropriate with white noise analysis to help in understanding the transformation of the sensory signals at the different neuronal stages. Morphological methods will reveal the shapes of individual neurons and the projections of their branches to particular regions of neuropil. Analyses will be made in an alert animal so that the actions of individual neurons can be related directly to behaviour. The intent is to use these favourable neurons and circuits to gain insights into specific features of sensory-motor processing and thereby generate ideas that are applicable to the design and functioning of local circuit neurons in more complex brains.